Refactor high-complexity React components in Dify frontend. Use when `pnpm analyze-component...
npx skills add digital-stoic-org/agent-skills --skill "edit-tool"
Install specific skill from multi-skill repository
# Description
Orchestrates creation of Claude Code tools (skills, commands, agents, scripts). Use when user requests creating, updating, or improving any Claude Code extension mechanism. Triggers include "create/make/new skill/command/agent/script", "tool for X", "slash command", "sub-agent", file paths with /skills/, /commands/, /agents/. Auto-triages based on token budget, frequency, context needs, and execution type. Delegates to edit-skill, edit-command, or edit-agent after explaining decision rationale.
# SKILL.md
name: edit-tool
description: Orchestrates creation of Claude Code tools (skills, commands, agents, scripts). Use when user requests creating, updating, or improving any Claude Code extension mechanism. Triggers include "create/make/new skill/command/agent/script", "tool for X", "slash command", "sub-agent", file paths with /skills/, /commands/, /agents/. Auto-triages based on token budget, frequency, context needs, and execution type. Delegates to edit-skill, edit-command, or edit-agent after explaining decision rationale.
Edit Tool Orchestrator
Automatically triages tool creation requests to the appropriate mechanism (skill, command, agent, or script).
Triage Process
CRITICAL: Always follow these steps:
- Analyze request against decision tree
- Explain decision to user with rationale
- Invoke appropriate skill or provide guidance
Decision Tree
graph TD
A[Tool Request] --> B{Deterministic<br/>shell only?}
B -->|Yes| C{AI needed to<br/>decide when/how?}
C -->|No| BASH[✅ Bash Script<br/>0 tokens]
C -->|Yes| D{Token<br/>budget?}
D -->|<500| WRAP[✅ Skill + scripts/<br/>AI wraps code]
D -->|>500| CMD1[✅ Command<br/>+ @filepath]
B -->|No| E{Token<br/>budget?}
E -->|>2000| AGENT[✅ Sub-Agent<br/>Isolated context]
E -->|500-2000| F{Invocation?}
F -->|User| CMD2[✅ Command<br/>User controls]
F -->|Auto| AGENT
E -->|<500| G{Frequency?}
G -->|5+ per session| H{Specific<br/>capability?}
H -->|Yes| SKILL[✅ Skill<br/>Auto-invoked]
H -->|No| CMD2
G -->|1-2 per session| CMD2
G -->|Rare <1/10| DIRECT[❌ Direct Request<br/>No tool needed]
style BASH fill:#90EE90,stroke:#000,stroke-width:2px,color:#000
style SKILL fill:#FFD700,stroke:#000,stroke-width:2px,color:#000
style WRAP fill:#FFA500,stroke:#000,stroke-width:2px,color:#000
style CMD1 fill:#87CEEB,stroke:#000,stroke-width:2px,color:#000
style CMD2 fill:#87CEEB,stroke:#000,stroke-width:2px,color:#000
style AGENT fill:#DDA0DD,stroke:#000,stroke-width:2px,color:#000
style DIRECT fill:#FFB6C1,stroke:#000,stroke-width:2px,color:#000
Decision Factors Reference
When path requires multiple criteria, consult:
| Factor | Key Question | Result |
|---|---|---|
| Token Budget | <500 / 500-2000 / >2000? | Skill / Command / Agent |
| Frequency | 5+ / 1-2 / <1 per 10 sessions? | Skill / Command / Direct |
| Context | Main / Isolated? | Command / Agent |
| Scripts | AI wrapper needed? | Skill+scripts/ / Bash |
| Invocation | User / Auto? | Command / Skill |
| Capability | Specific / Workflow? | Skill / Command |
Explanation Template
Explain decision: ✅ [TYPE] because: token budget (~X → range), frequency (pattern), key factor (dimension)
Example: ✅ SKILL because: ~300 tokens → <500, 10+/session, auto-invoked capability
Delegation
After explanation: Skill → edit-skill | Command → edit-command | Agent → edit-agent | Bash → Direct guidance (scripts/, #!/bin/bash, chmod +x, no tool)
Parallelization Check
Before any write operations, verify safety:
| Operation Type | Guidance |
|---|---|
| ✅ Read-only | Parallelize freely |
| ⚠️ Writes (independent files) | Sequential OR Plan Mode first |
| ⚠️ Writes (dependent changes) | Must sequence carefully |
| ❌ Destructive operations | Plan Mode MANDATORY |
Plan Mode triggers:
- Creating >3 files
- Modifying >5 files
- Architectural changes
- Breaking changes
Tool Comparison Quick Reference
| Tool | Token Cost | When to Use | Context |
|---|---|---|---|
| Bash Script | 0 (executed) | Deterministic shell ops | None |
| Skill | <500 ideal | Auto-invoked, frequent | Shared (pollution) |
| Command | 500-2000 ok | User-triggered workflows | User-controlled |
| Agent | Unlimited | Complex exploration | Isolated |
Common Patterns
| Pattern | When | Solution |
|---|---|---|
| Script Wrapper | AI decides timing for existing scripts | Skill + scripts/ |
| Verbose Workflow | >1000 tokens, manual trigger | Slash command |
| Research Task | Multi-file exploration, autonomous | Sub-agent (check Task tool first) |
See reference.md for edge cases, conversion guide, and extended examples.
# Supported AI Coding Agents
This skill is compatible with the SKILL.md standard and works with all major AI coding agents:
Learn more about the SKILL.md standard and how to use these skills with your preferred AI coding agent.